Humanitarian OSM Team/Working groups/Technical/meeting 2013-07-01
IRC meeting on Monday 1st July 2013
IRC log:
18:03 : '
18:03 mkl: greets
18:03 harry-wood: hello mkl
18:05 mkl: hey harry-wood, here for the tech chat?
18:05 harry-wood: yeah
18:05 ybon: hi there
18:05 harry-wood: mkl want me to blast a reminder to the mailing list? or you're doing it?
18:06 mkl: go for it
18:06 mkl: i've just arrived in nairobi, so a bit scattered
18:06 mkl: hi ybon
18:07 ybon: hi mkl :)
18:07 mkl: rbanick said he might be joining
18:08 mkl: even just us three is a great start
18:08 mkl: I don't know if we've ever had a regular technical forum for HOT
18:08 skorasaurus: also here for a bit.
18:09 : om [~omg@ma92436d0.tmodns.net] entered the room.
18:09 : vdeparday [~c05664cb@shenron.openstreetmap.org] entered the room.
18:09 flavour: here
18:09 om: Hey
18:09 mkl: there's so much development and possible ideas, i thought it would be worthwhile to just start regularly having tech focused chats
18:09 mkl: to help us coordinate, etc. but nothing too heavy
18:10 skorasaurus: k, so we're sharing we've done recently ?
18:10 mkl: similar to the idea of the Engineering Working Group with the OSMF, I suppose
18:10 : ian29 [~ian29@50-198-132-58-static.hfc.comcastbusiness.net] entered the room.
18:10 : pierregiraud [~pierregir@ver73-2-88-160-44-40.fbx.proxad.net] entered the room.
18:10 mkl: skorasaurus: i guess i was just doing intro. but good, that's a good starting place
18:10 mkl: what are we all up to?
18:11 mkl: maybe we can do a quick agile like check in
18:11 pierregiraud: is late
18:11 mkl: skorasaurus, want to kick us off?
18:11 skorasaurus: ok, sure.
18:11 : LS [~44a4ec54@shenron.openstreetmap.org] entered the room.
18:11 skorasaurus: just as a refresh, i was in northern haiti, as a part of the cap103 project.
18:12 skorasaurus: ybon was there from march-beginning of may,
18:12 harry-wood: You been working with ybon on the awesome map style stuff?
18:12 skorasaurus: i was beginning of may-present day
18:12 skorasaurus: harry-wood: yes.
18:12 skorasaurus: he deserves most of the credit :)
18:12 harry-wood: Everyone seems to love it
18:12 mkl: it's even being forked by unhcr
18:13 harry-wood: prompted lots of questions and quibbles on the mailing list
18:13 skorasaurus: we've done 2 mappy things, the josm-hdm preset and more importantly, the carto-css rendering.
18:13 skorasaurus: mkl: that's the tagging preset i believe.
18:13 AndrewBuck: yes, that rendering is a nice addition to the HOT toolset for sure.
18:13 pnorman: waves
18:13 skorasaurus: which was also done in our project, by us, led by brian wolford [wwolf]
18:14 mkl: https://github.com/unhcr/HDM-CartoCSS
18:14 skorasaurus: oh, awesome !
18:14 skorasaurus: didn't see that they forked that as well.
18:14 mkl: perfect, we are already ahead for this meeting ;)
18:15 AndrewBuck: 2 branches and they have 2 commits ahead on one as well, so not just forked but worked on.
18:15 skorasaurus: that's something we really had in mind, the ability to have ppl fork and customize for their specific needs.
18:15 mkl: right on
18:15 AndrewBuck: Yeah, I assume the CSS is more amenable to that than the old XML styles were.
18:15 skorasaurus: and we'll incorporate those back into ours as needed.
18:15 skorasaurus: oh yes.
18:15 ybon: oh, didn't noticed this fork, grat
18:16 ybon: great
18:16 mkl: how is this style being deployed?
18:16 skorasaurus: hosted ?
18:16 mkl: right
18:16 skorasaurus: right now, it's on ybon's server.
18:16 ybon: by the way, I miss some notifications for the hotosm repository, I don't know if someone can do something for this
18:16 ybon: mkl: on my dev server for now
18:16 AndrewBuck: I had heard some murmurings that you were looking for a better server to host it in the future, correct?
18:16 ybon: and working right now on a wordwild deployement
18:16 skorasaurus: AndrewBuck: yes.
18:16 ybon: AndrewBuck: I'm on it
18:17 AndrewBuck: Does HOT have a server with some power?
18:17 skorasaurus: that's something this group can help develop.
18:17 mkl: cool. i suppose this is one thing the sysadmin, once hired, can start to help with
18:17 ybon: on the OSM-fr servers
18:17 AndrewBuck: ybon: ahh, good idea.
18:17 skorasaurus: oh cool, ybon i wasn't sure if that was going to be on osm-fr or not.
18:17 skorasaurus: yeah, mkl, regarding the sys-admin.
18:17 mkl: note to us, we should at least work on a catalog of all the hot services, and where they're hosted, and who's responsible
18:17 skorasaurus: i haven't heard anything regarding that since we put out the call.
18:18 rbanick: hey i'm here, belatedly
18:18 mkl: sure, i can update on that
18:18 mkl: we are scheduling interviews with the finalist candidates
18:18 : dalekunce [~a2066112@shenron.openstreetmap.org] entered the room.
18:18 skorasaurus: mkl: yeah, one big takeaway that I had from my time in haiti was the need, around HOT is better documentation.
18:18 mkl: if everything moves as expected, we could have a sysadmin in 2 weeks
18:18 skorasaurus: mkl: great, was about to ask that.
18:18 mkl: should we just start a wiki page now?
18:19 skorasaurus: why not.
18:19 skorasaurus: I've started tweaking the main HOT wiki page as well.
18:19 mkl: great, that's really needed
18:19 om: need fresh eyes though, too easy to assume reader has a certain level of understanding
18:20 mkl: anyone else want to give a check in?
18:20 mkl: ybon?
18:20 skorasaurus: om, true.
18:20 AndrewBuck: On the subject of wiki tweaking, the Activation Group decided it would be good to come up with some templates to standardise wiki pages for activation and pre-activation pages as well.
18:20 harry-wood: ybon You working on getting it working worldwide on your dev server?
18:20 AndrewBuck: Just thought I would through that out here as it is somewhat related.
18:20 ybon: mkl: sure
18:20 mkl: AndrewBuck: indeed, long overdue
18:20 ybon: harry-wood: nope, on the OSM-fr servers
18:20 harry-wood: aha
18:20 ybon: we have all the chain with minutes diff
18:21 ybon: the thing is that we have also update it this week-end, so it takes time to regenerate every metatile for every style...
18:21 AndrewBuck: ybon: does that rendering use the same DB as the normal mapnik one?
18:21 ybon: AndrewBuck: yes
18:21 AndrewBuck: cool
18:21 ybon: this was one of my constraint
18:21 pnorman: requires hstore?
18:21 ybon: to be able to host it on every shared server
18:21 ybon: pnorman: yes
18:21 mkl: btw, we have this page for Tech http://wiki.openstreetmap.org/wiki/Humanitarian_OSM_Team/Working_groups/Technical
18:21 pnorman: not much of a constraint then :)
18:22 AndrewBuck: pnorman: true, but better than some custom schema entirely. :)
18:22 ybon: pnorman: who doesn't use hstore in osm2pgsql nowafays? ;)
18:22 skorasaurus: ybon: you'll have to implement slim mode to do diffs though, right
18:22 pnorman: orm doesn't have hstore
18:22 ybon: I will not, diff are already working
18:22 skorasaurus: oh, ok, cool. good to know.
18:23 ybon: pnorman: yeah but I think everyone else does, and how do you use tags not in the default osm2pgsql style without changing it and without hstore? ;)
18:23 mkl: how's umapper going?
18:23 pnorman: you'll have to use slim mode to import the planet unless you have hundreds of GB of ram, but it's just a command-line flag
18:23 ybon: mkl: uMap? :)
18:23 mkl: i need to switch to using hstore for my tiles!
18:23 mkl: ybon: uh, yes, that one ;)
18:24 ybon: ok :)
18:24 ybon: so it's going well
18:24 ybon: just rolled out the O.4
18:24 ybon: with some improvment, like the data browser
18:24 ybon: and better IE support
18:24 AndrewBuck: ybon: Great work on that. I really like the forward backward arrows in the popups for objects so you can jump to the next without zooming out. More of that kind of thing is really cool.
18:25 ybon: AndrewBuck: cool, thanks
18:25 ybon: taking this into account for future improvement, so :)
18:25 AndrewBuck: It is not just the standard map control widget everyone makes, a really innovative approach.
18:26 mkl: anything more ybon?
18:26 ybon: humm
18:26 mkl: we're still doing go rounds
18:26 ybon: I think it's ok for me then
18:27 ybon: just adding that I were two months in Haiti two in march and April this year in case someone doesn't know it
18:27 mkl: and thank you for that, in case you haven't heard it enough!
18:28 AndrewBuck: ybon: was it you that set up the caching server there for imagery?
18:28 ybon: yes
18:28 mkl: who's next?
18:28 AndrewBuck: ybon: that is something we should make use of for the other field teams. I assume the ones in africa and elsewhere will have similar issues.
18:29 ybon: yeah, sure, but having a tech on site is something, not having one, another ;)
18:29 mkl: where was the imagery cached?
18:29 mkl: on a server in haiti? or a laptop?
18:29 ybon: a desktop Ubuntu
18:29 ybon: with a Squid
18:29 ybon: on the univ
18:29 AndrewBuck: I don't know the technical stuff about squid or whatever you used but if you want to document it and need a set of eyes for proofreading the docs and whatnot I would be happy to help with what I can.
18:29 ybon: used as proxy
18:30 ybon: AndrewBuck: draft http://wiki.openstreetmap.org/wiki/User:Ybon/Squid
18:30 ybon: not really a doc, but just not to forget what was done
18:30 AndrewBuck: ybon: perfect. Will read through it and see if I can add anything then.
18:30 AndrewBuck: I might try replicating it here in a VM or something.
18:30 ybon: oh it's just a memom ;)
18:30 ybon: memo
18:30 AndrewBuck: Then I can learn it and at the same time we might get a useful VM we can just ship out to the teams.
18:30 mkl: there's the documentation need again
18:31 AndrewBuck: No admin necessary then.
18:31 mkl: nice idea
18:31 ybon: the problem is that servers work until there is sysadmin eyes around ;)
18:31 harry-wood: On that topic... there once was a plan to develop a usb key of HOT awesomeness.
18:31 ybon: as soon as one take it's plane back to France, so the server know it, and start doing jokes ;)
18:32 AndrewBuck: harry-wood: is this like an OS on a usb?
18:32 AndrewBuck: boot into a HOT enviro, etc?
18:32 mkl: what awesomeness would be there?
18:32 om: that sounds like a winner
18:32 harry-wood: Maybe an OS... or just loads of software and data
18:32 harry-wood: assuming an OS
18:32 ybon: harry-wood: we have done a custom Ubuntu image in Haiti
18:32 skorasaurus: harry-wood: was that the HOT-package ?
18:32 ybon: with JOSM, QGIS and so on
18:33 ybon: but as soon as you do the image, it is already not up to date
18:33 ybon: so I think we need more recipes than images
18:33 harry-wood: yeah could be
18:33 AndrewBuck: harry-wood: we have been kicking around the idea of making a VM image with walking papers server, a tile server, an overpass server fed by minutelies, etc as well as things like this squid cache, etc, to be deployed in such a fashion.
18:34 AndrewBuck: ybon: yeah agree. I have heard good things about vagrant, but only just started learning that. Would be a good choice for something of that nature though I think.
18:34 ybon: (in a raspeberripy, of course)
18:34 mkl: i'm excited to see what the interface for http://brck.com/ is going to be like
18:34 pierzen: The basic idea is to try remove as much as possible from the poor internet bandwith.
18:34 harry-wood: mkl: ah yes
18:34 harry-wood: so I was talking to the ushahidi devs
18:34 harry-wood: a couple of days ago
18:34 ybon: ah intereseting
18:34 AndrewBuck: mkl: yeah, this server would pair nicely as pierzen mentioned.
18:34 harry-wood: and I got them excited about the idea of bundling some OSM data on there
18:34 ybon: nice
18:34 pierzen: harry-wood: this is only taking care of connecting to internet.
18:34 mkl: right on
18:35 harry-wood: pierzen no that's what I thought, but apparently it'll do a bit of storage too
18:35 mkl: it's actually supposed to do more, though they're focusing on that to start
18:35 mkl: ie, it can provide services locally
18:35 AndrewBuck: mkl: does brck actually have an OS like linux in it on its own, or is it only for serving internet connectivity.
18:35 mkl: i've gotten different answers to that question myself
18:35 ybon: ( AndrewBuck http://www.docker.io/ also )
18:36 harry-wood: not sure about OS. It's presumably more than just an aduino board
18:36 harry-wood: ...but maybe not much more
18:36 mkl: harry-wood: who are you talking to? i'm just now 2 minutes from the iHub
18:36 AndrewBuck: ybon: that is like vagrant then?
18:36 harry-wood: mkl I was talking to a couple of developers who were meeting in London
18:36 harry-wood: mkl David Kobia
18:37 mkl: k
18:37 harry-wood: and Seth something
18:37 ybon: AndrewBuck: same familly
18:38 ybon: AndrewBuck: docker is more on bundling, vagrant more on configuration/management/deployment
18:38 harry-wood: I thought dane springmeyer had done some work on a USB key of awesomeness
18:38 mkl: keeping the go rounds, harry-wood do you have anything to update us all on?
18:38 harry-wood: but maybe it's this thing I'm thinking of https://github.com/hotosm/installer
18:38 AndrewBuck: ybon: ok. Will check it out then
18:38 mkl: we can keep on discussion on various points, just wanted to start out by us getting a sense of everything going on
18:39 harry-wood: mkl: update from me: Very busy at the moment. Trying to keep plates spinning, so not really delving deep into any tech. I'll be working with schyler on interview a sysadmin
18:39 mkl: cool
18:40 harry-wood: oh and I'll be looking over what frederik's been doing developing hot-exports
18:40 harry-wood: (geofabrik paid dev work)
18:40 mkl: harry-wood: so you are doing oversight on that contract for HOT?
18:41 harry-wood: well... I'm not sure if I'm in charge of oversight, but was assuming I'm expected to look over that work
18:41 harry-wood: brian woolford has also been looking over it I think
18:41 pierzen: yes and myself
18:42 harry-wood: Anyway latest update from Frederik was that he checked everything into github and is looking over the issues logged there
18:42 mkl: cool. has something been delivered yet for hot-exports?
18:42 mkl: ah, ok
18:42 harry-wood: We are... somewhere... on the delivery roadmap yes
18:42 mkl: anyone else want to jump in? pierzen, AndrewBuck, rbanick?
18:43 mkl: AndrewBuck: saw a bunch of emails from you about tasking server, etc. (i haven't read everything myself yet) anything you're looking into in particular?
18:43 skorasaurus: harry-wood: he's delivered an unfinished version of it.
18:43 AndrewBuck: I would like to raise an issue about getting a mumble server set up for HOT.
18:44 skorasaurus: but unfortunately ,he hadn't been very responsive to issues that we raised.
18:44 ybon: +1 for the Mumble server
18:44 AndrewBuck: mkl: no not on the tasking server. Just kincking in my 2 cents
18:44 dalekunce: This is Dale, I'll speak on behalf of ARC
18:44 mkl: what is Mumble?
18:44 pierregiraud: I also read a lot of things about the tasking manager
18:44 AndrewBuck: There is some discussion about making a more micro tasking system than the current tasking server and I just contributed my perspective from what went on in Mali with the Kuona tasking platform.
18:44 harry-wood: skorasaurus I think frederik isn't really of the highly-responsive-on-github developer new-school
18:45 AndrewBuck: Mumble is like teamspeak but opensource.
18:45 dalekunce: We've had a bunch of talk recently internally and are really excited about a lot of things this group is doing, ie, HOT on a stick
18:45 AndrewBuck: It is basically like an IRC server except you voice chat instead of type. (it has text chat as well)>
18:45 skorasaurus: harry-wood, i just saw fred updated the github about 3 days ago :)
18:45 mkl: AndrewBuck: cool, microtasking is a good discussion to have. lots of tools out there
18:45 skorasaurus: i haven't used it in a week or 2 [the exports)
18:46 AndrewBuck: mkl: we would need a server with a permanent hostname and/or IP address and a reasonable internet connection for mumble.
18:46 AndrewBuck: Does HOT have such a thing?
18:46 AndrewBuck: Bandwidth needs are not too overly high but you want low latency, etc.
18:46 mkl: we don't know what we have exactly
18:46 AndrewBuck: :)
18:46 harry-wood: we can stick things on the OSM dev server within reason
18:46 harry-wood: not sure about a mumble server though
18:46 AndrewBuck: ybon: could something like that be put on the osm.fr server potentially?
18:47 AndrewBuck: ybon: might be asking too much though with the already request for your tile server.
18:47 mkl: dalekunce: did you want to say more about what the ARC is thinking?
18:47 AndrewBuck: That is more important.
18:47 AndrewBuck: harry-wood: I asked about running it on the dev server but it sounds like the firewall would block it (and UCL controls that, not us).
18:48 ybon: AndrewBuck: at the moment, we are on top of what we can provide
18:48 AndrewBuck: ybon: Yeah, I figured that.
18:48 dalekunce: ARC has given a lot of though about what/how we can best contribute to this community
18:48 AndrewBuck: I am runnning mumble on my home PC right now, but my IP address changes periodically.
18:48 AndrewBuck: oh, Hi dale. I miseed you joining. Welcome.
18:49 dalekunce: given money is always a problem for everyone, we thought the next best thing we could do was to organize something
18:49 AndrewBuck: Dale and Robert Bannick were discussing the HOT vm with us.
18:49 dalekunce: We are thinking about and have recieved internal approvals to host a HOT codesprint to support a lot of this topics
18:49 AndrewBuck: dalekunce: that was mentioned briefly above, not sure if you saw it or not.
18:50 AndrewBuck: A codesprint would be awesome.
18:50 mkl: i had been thinking of a code sprint for OSM's birthday, next month
18:50 dalekunce: joining redcross and hot of course gives us the awesome name, RedHOT
18:50 AndrewBuck: :)
18:50 mkl: dalekunch: whoa, brilliant
18:51 mkl: or we could do world humanitarian day again
18:51 harry-wood: on the topic of HOT usb stick.... https://vine.co/v/b0DvTPnpPtw <<< The entire word rendering on the fly!
18:51 harry-wood: We should get this kind of technology onto brcks and into the hands of aid workers and disaster responders
18:51 pierregiraud: talking about a code sprint, what's the status of the hot week in maptember?
18:51 dalekunce: the mobile VM/server/stick is a priority for us
18:52 : f00bar left the room (quit: Ping timeout: 480 seconds).
18:52 : xivk [~4e150242@shenron.openstreetmap.org] entered the room.
18:52 mkl: harry-wood: do you know the details on the hot week?
18:52 dalekunce: harrywood: thats our thoughts exactly
18:52 harry-wood: HOT week? the event after SOTM you mean?
18:52 harry-wood: I think we were hoping to attract sponsors for it.... and failed
18:53 mkl: I do think we're trying to schedule a face to face board meeting then
18:53 ybon: harry-wood: no POI at all in this on the fly demo
18:53 rbanick: hey guys, got called away...just joining up
18:53 rbanick: I see the RedHot idea is getting aired.
18:53 harry-wood: ybon yeah. I guess that's a limitation of their early prototype there.
18:53 dalekunce: we have teams that go to disasters hours/days after and it would be awesome if the EOC had HOT support
18:53 pierregiraud: harry-wood: is it canceled then?
18:54 harry-wood: pierregiraud: I think so. Unless organisational miracles happen
18:54 rbanick: yeah the HOT USB stick would be really helpful for us on larger deployments where we had GIS capable staff.
18:54 dalekunce: we are happy to fit into the calendar wherever
18:54 rbanick: or if it was simple enough to train our not tech brilliant responders...
18:54 rbanick: si, we're flexible.
18:55 ybon: Is there already some brainstorming wiki page about this HOT USB thing?
18:55 ybon: just to be sure of what everyone has in mind, in terms of usage
18:55 dalekunce: to support the HOT USB we also want to build up uMap as well, so non GIS folks can contribute in real time as well
18:55 harry-wood: ybon : I thought there was. Vaguely remember dane talking about it... but not sure where
18:56 mkl: hey y'all, i'm going to check out in 5 minutes
18:56 harry-wood: Of course the other product with similar awesomeness, which is already out there... is MapsWithMe.
18:56 mkl: wanted to know:
18:56 mkl: was this useful? next one when?
18:57 mkl: anything for us to work on coming out from this discussion?
18:57 harry-wood: in a few weeks we may have hired a sysadmin who may find it useful to gather ideas with a session like this
18:57 AndrewBuck: mkl: I would say it has been useful
18:57 om: task manager was to simplify the process for unfamiliar folks.. there is also room for crisismap-like interface that could store annd manage a lot of useful crisismap info. for deployments
18:58 : rbanick left the room (quit: Remote host closed the connection).
18:58 om: (SBTF and the like)
18:58 harry-wood: om yeah. I guess umap fits into that need a bit
18:58 harry-wood: and everything ushahidi does of course
18:58 mkl: i'd like to gather up some simple lists of HOT tech projects
18:59 mkl: just a listing of what's active, where things are being hosted, and what's needed
18:59 mkl: i would start that on the wiki, but I've been locked out for too many bad password attempts :(
18:59 pierzen: Contributor statistics : I have improved Sev Pierrel Python script that produce statistics interrogating OSM through OSMApi for a period of time and a bounding box.
19:00 pierzen: We provide arrays of contributors by team, the period to cover and a bounding box. Statistics of objects created, modified or deleted (nodes, ways and relations). Some POI's statistics also. Output has csv files.
19:00 pierzen: I want also to develop a less sophisticated version that would provide net statistic comparing two planet files. Then, we would only have net objects created. But we could cover a larger geographic zone.
19:00 om: needed: quickstart map page for crisis response and docs about using OSM as baselayer.
19:00 AndrewBuck: om is this for the ushahidi platform?
19:01 pierzen: We need a clear workflow with other NGO to assure they use appropriate tools that let them use OSM base layer.
19:02 AndrewBuck: pierzen: +1
19:02 harry-wood: appropriate tools?
19:02 pierzen: other then google crisis map.
19:02 pierzen: for example.
19:03 ybon: maybe a umap.hotosm.org? :)
19:03 AndrewBuck: harry-wood: lots of groups have been using google maps as their base layer so we can't reintegrate field data into OSM from the google heritage of the geolocation...
19:03 pierzen: Crowdmap , Umap or similar tools lets show OSM base layer. :)
19:03 AndrewBuck: This happened both in india and amadagascar.
19:03 AndrewBuck: *madagascar
19:03 AndrewBuck: We lost a lot of field data in madagascar to that issue.
19:03 harry-wood: Yeah that's annoying
19:03 pierzen: But in Madagascar, they used Crowdmap and the OSM base layer was used.
19:04 AndrewBuck: They had field teams marking hospitals, etc, on crisis map but it was all with google as background so we had to just leave it.
19:04 AndrewBuck: pierzen: much was on google.
19:04 AndrewBuck: They did use OSM some.
19:04 mkl: http://wiki.openstreetmap.org/wiki/Humanitarian_OSM_Team/Working_groups/Technical
19:04 harry-wood: All we need to do is persuade the entire developer community that using google maps is uncool
19:04 AndrewBuck: harry-wood: +1
19:04 harry-wood: should be easy
19:04 skorasaurus: harry-wood: most ;)
19:04 skorasaurus: a few already are. :)
19:04 mkl: I just started to add some sections, based on the discussion today. Please take a minute and fill in.
19:05 mkl: and can someone post the chat log here as meeting minutes?
19:05 harry-wood: mkl I'll do that
19:05 mkl: cheers harry-wood
19:05 pierzen: More then that, we should have prerequisite when we start to collaborate. If we see they make no effort to use OSM, we should stop collaborating on a specific mapping project.
19:05 harry-wood: (provided the meeting wraps up soon)
19:06 ybon: btw, I was thinking to suggest to blog this on the chan: https://botbot.me/ maybe, thoughts?
19:06 om: Having been on the 'other' side, I can tell you the decisions are never deliberate to not use OSM, it's usually the initiator picked it out of convenience and familiarity
19:06 ybon: sorry, to plug* not to blog*
19:06 harry-wood: I think it's important not to alienate people by being very OSM-purist
19:06 harry-wood: but sadly the issue does mean we have to turn down data which otherwise we might've used
19:06 AndrewBuck: yes, the activation working group discussed this as a key part of formal activation, that we _must_ have indications the data will be used before we will formally activate on aresponse.
19:07 pierzen: it is not purist. If they ask us to collaborate, we expect that they will use osm data.
19:07 skorasaurus: otherwise, they could go to crisismappers.
19:07 AndrewBuck: om: that is what we want to work on though. We want NGO's to know what we can do and what is available so they do choose us.
19:07 om: I believe that if we put a little attention into the actual info needs of crisismapping groups and presented them an interfacee with convenient features it would get used. Even a primer on how/why to use OSM baselayer would be valuable
19:08 om: then what happens is people think it means switching tools midd-stride and they freak out.
19:08 AndrewBuck: om: yeah.
19:08 om: we can ease this anxiety easily
19:09 AndrewBuck: om: the other thing that needs to happen is these outside groups need to interface us _before_ disasters hit so when they do respond they are already familiar with our system.
19:09 mkl: which groups are we talking about here?
19:09 AndrewBuck: om: if you have contcts at NGO's we would be happy to talk to them to let them know why OSM is a good choice.
19:09 pierzen: That is why we need a clear workflow. One is asking our help. A second person decide to use an incompatible tool...
19:10 mkl: i know with India, they were concerned to swtich because they were working with Google
19:10 om: I agree and have been trying to help that process within SBTF, bbut it will be better I think if HOT puts out a 'welcome mat'
19:10 pierzen: it is not important to discuss about specific groups on this list.
19:10 pierzen: It can happen with any group.
19:10 AndrewBuck: mkl: were they working with google specifically or just using the google map?
19:10 mkl: they had support of g crisis response
19:11 AndrewBuck: mkl: ahh, I see. That makes sense for that case I guess.
19:11 om: mkl: it wasn't a deliberate or thought out decision
19:11 mkl: pierzen: sure, true. i just know that we do have relationships w/ many groups, and curious to know what's missing
19:11 AndrewBuck: They still should be aware of us though.
19:11 pierzen: Then why did they asked us to collaborate and liaison on skype?
19:11 : laurentS [~laurent@c80-216-218-81.bredband.comhem.se] entered the room.
19:11 mkl: pierzen: exactly! was pretty confusing
19:12 om: it was just their interface and that others could add layers as kmml (knoesis had kml chops)
19:12 mkl: back to your comment about activation protocols and clear partners
19:12 pierzen: or clear procedures.
19:12 AndrewBuck: om: the other thing that is missing, is not just the using of ourr map, but getting feedback from the NGO so we know what/where to map.
19:12 mkl: om: i actually implemented OSM baselayer in that tool, when it was released open source. had help from ping too. so it wasn't technical
19:12 AndrewBuck: We can't just get a request like "map india" and then go to work.
19:12 mkl: to be clear about India, I only floated the idea to the local Indian community to take forward
19:13 AndrewBuck: We need to know specific towns, what are the priorities (roads or buildings, etc) and when towns stop being a priority (after they have been helped, etc).
19:13 mkl: i think the action just grew beyond the discussion
19:13 om: with a more familiar interface, they would have used OSM, and iif they knew how to switch baselayers would have also. It did eend pretty quickly and the mapping wasn't major part. But it ought to be! Kitemapping is probably a great healing activity as well as being very useful for recovery
19:13 AndrewBuck: So it needs to be a whole relationship with the NGO we work with, not just a one off thing..
19:14 pierzen: This brings this idea of having better liaison between events and plan how we will interface. And in general, they already know very well Crowdmap.
19:14 AndrewBuck: mkl: yeah, It was hard to tell with india how far the activation was meant to go, or even if we did formally activate or not.
19:14 AndrewBuck: mkl: that is something the activation group discussed.
19:14 mkl: cool
19:15 AndrewBuck: mkl: we are trying to be more clear in the future with guidelines for activations, etc.
19:15 mkl: yup, it's overdue
19:15 pierzen: we are to be flexible, to start rapidly, establish contacts and evaluate situation.
19:15 mkl: when's the next Activatoin WG meeting?
19:15 pierzen: And eventually decide to continue or stop.
19:15 mkl: yea, we never actually stop
19:15 AndrewBuck: mkl: om: see the bullet points at the top of the log from the last meeting about that specifically...
19:15 AndrewBuck: http://wiki.openstreetmap.org/wiki/Humanitarian_OSM_Team/Working_groups/Activation/meeting_2013-06-26
19:15 pierzen: not yet decided. Had a very good 3 hours discussion.
19:15 om: from the earliest days crisis mapping takes the path of least chaallenge (gdocs and skype with ushahidi) rather than create new capacity. HOT is opposite (in a good way)
19:16 AndrewBuck: om: One thing that would help a lot is feedback on the mailing list from people requesting data from us.
19:17 mkl: well for this meeting, i'm stepping out, but feel free to keep going on the tech topic, this is awesome
19:17 mkl: harry-wood, log when you think appropriate
19:17 mkl: everyone, fill in the wiki http://wiki.openstreetmap.org/wiki/Humanitarian_OSM_Team/Working_groups/Technical
19:17 AndrewBuck: If an NGO is active and want our help they should post to the mailing list every 12 to 24 hours a brief note with status, priorities, etc, so we don't mis-direct our efforts.
19:17 mkl: and maybe we do it all again in 2 weeks
19:17 harry-wood: om: The desire for good documentation on how to use OSM is very much not limited to humanitarian work. HOT did attack this problem a bit with learnosm.org ....but that doesn't cover deploying slippy maps or "using" in general
19:18 harry-wood: maybe it should
19:18 pnorman: I was talking to RichardF about switch2osm
19:18 harry-wood: yeah switch2osm.org is on that topic
19:18 AndrewBuck: harry-wood: I think documentation is less the issue, than just curating a lasting relationship with these groups. We tend to focus too much on the activations, and not eneough between activations.
19:19 harry-wood: sure
19:19 harry-wood: that reminds me I really must meet up my London MSF contact again some time
19:19 om: well, another issue is that mappers are using what they know about or what gets promoted through the community = crowdmmap and ushahidi and now google crisis map. They have needed a purpose built tool for ages.
19:19 mkl: AndrewBuck: to be fair, a good amount of relationship building all the time. it's just not as visible and coordinated as it should be
19:19 AndrewBuck: mkl: that is true.
19:20 AndrewBuck: I forget that there is work like that going on.
19:20 AndrewBuck: I know it is slow. :)
19:20 om: I also like creating a 'crisis layer' in wikimapia that is distinct from their typical closed license API
19:20 AndrewBuck: om: we can't have anything at all to do with wikimapia.
19:20 harry-wood: eugh wikimapia
19:20 pnorman: the tile server guides need a rewrite for carto, so there's work that needs doing on s2o anyways, I think s2o needs a bit more direction and a bit clearer of a split between switching websites and rendering tiles
19:20 AndrewBuck: they are toxic
19:21 mkl: a clear guide to what data sources we can use, and what we can't , and why, and how to change that ... that would be useful
19:21 AndrewBuck: om: the problems with wikimapia go way beyond the license. I would avoid them at all costs unless you want to get drug into a very costly legal battle.
19:21 mkl: expalinging again and again why we can't use gmaps satellite is not so fun
19:22 AndrewBuck: mkl: that is a good idea.
19:23 harry-wood: I am going to have to go unfortunately
19:23 AndrewBuck: it sounds like we are getting near the end anyway.
19:24 harry-wood: Oh ok
19:24 skorasaurus: one last thing.
19:24 harry-wood: a closing tech topic?
19:24 skorasaurus: IOM Drone imagery. I'm not on the board, but I'd just like to find out what's HOT's stance on it :)
19:25 : hjart [~hjart@37.128.215.244] entered the room.
19:25 harry-wood: We had a few discussion on the topic
19:25 harry-wood: It's a tricky one organisationally, mainly issue around liability
19:25 mkl: yea, we want to work towards a policy. nicolas was tasked with gathering questions on UAV use
19:26 skorasaurus: k, didn't know if there was any recent discussions on it.
19:26 pnorman: anyways, https://wiki.openstreetmap.org/wiki/WikiProject_Haiti/ImageryAndDataSources#IOM needs a rewrite. "IOM is outdated" "IOM is the most current imagery"
19:26 skorasaurus: pnorman: will do that ^
19:26 mkl: need to publish the minutes
19:27 harry-wood: I wouldn't discourage anyone from experimenting with such things, but we need to be cautious about launching UAVs in the name of HOT
19:27 harry-wood: But yeah I think there's still more board discussion to be had on that topic
19:28 mkl: i just asked joseph to publish the last board minutes
19:28 mkl: ok bye all ... we drifted from tech, but the specific and general hot discussion all very useful
19:28 mkl: bye
19:29 harry-wood: so let's do another tech meeting in about 1 months time
19:29 harry-wood: hopefully after we've made progress with hiring a sysadmin
19:29 harry-wood: sounds good?
19:29 AndrewBuck: Yeah that seems reasonable.
19:29 ybon: AndrewBuck: liaison is in 30min?
19:30 AndrewBuck: no it is later than that I think, a few hours.
19:30 AndrewBuck: I will skype you the details.
19:30 harry-wood: ok I have to head off then
19:30 ybon: ok
19:30 ybon: got to eat so
19:30 harry-wood: I shall declare the meeting closed